Method and device for predicting operation state of building new energy system

Abstract

The invention discloses an operation state prediction method and device of a building new energy system, and the method comprises the steps: obtaining various monitoring variables of a target building, inputting the various monitoring variables of the target building into a trained prediction model, and obtaining the prediction results of the various monitoring variables of the target building in a set prediction time range; based on the prediction result, evaluating the operation state of the target building: firstly calculating a basic threshold value, and then determining a dynamic threshold value according to the basic threshold value, an equipment aging factor and a seasonal factor; the prediction result is compared with a dynamic threshold value, and if the prediction result is smaller than or equal to the dynamic threshold value, the operation state of the building new energy system is a normal state; and if the prediction result is greater than the dynamic threshold value, indicating that the operation state of the building new energy system deviates from the normal interval and the building new energy system has an abnormal phenomenon.

Description

Technical Field

[0001] This invention relates to the field of operational status prediction technology, and in particular to a method and apparatus for predicting the operational status of building renewable energy systems. Background Technology

[0002] The large-scale application of renewable energy sources such as building-integrated photovoltaics (BIPV), solar thermal energy (CTE), and heat pumps has become a significant demand. Due to the randomness, intermittency, and uncertainty of renewable energy, the operating status of building renewable energy systems is affected by various factors such as weather conditions, load demand, and equipment performance. Accurate prediction and assessment of the operating status of renewable energy systems is one of the key technologies to ensure their efficient operation. Existing methods for identifying deviations in the operating status of building renewable energy systems mostly involve comparison with fixed normal operating ranges or analysis based on various fault experience data. In methods that set fixed normal operating ranges, to avoid frequent state deviation alarms, the normal range setting is often quite broad. This leads to the inability to promptly and accurately identify slight decreases in the operating efficiency of the renewable energy system or when it is in an unhealthy state, resulting in long-term inefficient operation. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides a method and apparatus for predicting the operational status of building renewable energy systems; it can effectively improve the prediction accuracy of building renewable energy systems and significantly increase the utilization efficiency of renewable energy systems.

[0004] On the one hand, it provides a method for predicting the operational status of building new energy systems; Methods for predicting the operational status of building renewable energy systems include: Acquire historical monitoring sample data of building new energy systems and preprocess the historical monitoring sample data; A prediction model is constructed, and the prediction model is trained using preprocessed historical monitoring sample data to obtain the trained prediction model. Obtain the various monitoring variables of the target building, input the various monitoring variables of the target building into the trained prediction model, and obtain the prediction results of the various monitoring variables of the target building within the set prediction time range; Based on the prediction results, the operating status of the target building is evaluated: first, a basic threshold is calculated, and then a dynamic threshold is determined based on the basic threshold, equipment aging factors, and seasonal factors; the prediction results are compared with the dynamic threshold. If the prediction results are less than or equal to the dynamic threshold, it indicates that the operating status of the building's new energy system is normal; if the prediction results are greater than the dynamic threshold, it indicates that the operating status of the building's new energy system deviates from the normal range, and there is an abnormal phenomenon in the building's new energy system.

[0005] On the other hand, it provides a device for predicting the operational status of building new energy systems; The building renewable energy system operation status prediction device includes: The acquisition module is configured to: acquire historical monitoring sample data of the building's new energy system and preprocess the historical monitoring sample data; The training module is configured to: build a prediction model, train the prediction model using preprocessed historical monitoring sample data, and obtain the trained prediction model. The prediction module is configured to: acquire various monitoring variables of the target building, input the various monitoring variables of the target building into the trained prediction model, and obtain the prediction results of the various monitoring variables of the target building within the set prediction time range; The status assessment module is configured to: assess the operational status of the target building based on the prediction results; first, calculate the basic threshold, and then determine the dynamic threshold based on the basic threshold, equipment aging factors, and seasonal factors; compare the prediction results with the dynamic thresholds; if the prediction results are less than or equal to the dynamic thresholds, it indicates that the building's new energy system is operating normally; if the prediction results are greater than the dynamic thresholds, it indicates that the building's new energy system is deviating from the normal range and that there are abnormal phenomena in the building's new energy system.

[0006] Furthermore, an electronic device is also provided, including: Memory, used for non-transitory storage of computer-readable instructions; and Processor, for executing the computer-readable instructions, When the computer-readable instructions are executed by the processor, they perform the method described in the first aspect above.

[0007] In another aspect, a storage medium is also provided for non-transitory storage of computer-readable instructions, wherein when the non-transitory computer-readable instructions are executed by a computer, the method described in the first aspect is performed.

[0008] In another aspect, a computer program product is also provided, including a computer program that, when run on one or more processors, is used to implement the method described in the first aspect above.

[0009] The above technical solution has the following advantages or beneficial effects: This invention acquires historical, fine-grained time-series monitoring sample data of building renewable energy systems such as solar photovoltaic systems, solar thermal systems, and ground source heat pump systems, as well as building sub-metering data, and performs outlier processing and missing value imputation on the sample data. It employs a Transformer model based on an attention mechanism to construct and train a multivariate medium- to long-term prediction model for the building renewable energy system. The trained Transformer model is then used to synchronously predict various monitoring variables of the target building, generating predicted values ​​for a future period through a decoder. The MAPE (Mean Absolute Percentage Error) value is used to calculate the periodic deviation between the synchronous prediction results and the actual operational monitoring results, assessing the operational status of the building renewable energy system. When the renewable energy system is not in its optimal operating state, timely adjustments and optimizations are made to improve the utilization rate of renewable energy. The Transformer model performs excellently in tasks such as massive data processing, data mining, and machine learning. Its greatest advantage is its ability to perform parallel computation, improving processing speed and efficiency. Attached Figure Description

[0010] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0011] Figure 1 This is a flowchart of the method in Example 1. Detailed Implementation

[0012] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0013] Meteorological conditions, load demand, and equipment performance not only affect the operating status of specific building renewable energy systems (such as photovoltaic systems), but also other renewable energy systems within the building (such as solar thermal systems and heat pump systems). By simultaneously considering the changing patterns of other renewable energy systems in the prediction of the status of a specific renewable energy system, the feature vector space of the prediction variables can be effectively expanded. Therefore, by introducing the attention mechanism of the transformer algorithm, the variables among different renewable energy systems in a building can be considered simultaneously, thereby improving the accuracy of variable prediction.

[0014] Example 1 This embodiment provides a method for predicting the operational status of a building's new energy system; like Figure 1 As shown, the method for predicting the operational status of a building's new energy system includes: S101: Obtain historical monitoring sample data of building new energy systems and preprocess the historical monitoring sample data; S102: Construct a prediction model, train the prediction model using preprocessed historical monitoring sample data, and obtain the trained prediction model; S103: Obtain the various monitoring variables of the target building, input the various monitoring variables of the target building into the trained prediction model, and obtain the prediction results of the various monitoring variables of the target building within the set prediction time range; S104: Based on the prediction results, assess the operational status of the target building: first calculate the basic threshold, then determine the dynamic threshold based on the basic threshold, equipment aging factors, and seasonal factors; compare the prediction results with the dynamic threshold. If the prediction results are less than or equal to the dynamic threshold, it indicates that the building's new energy system is operating normally; if the prediction results are greater than the dynamic threshold, it indicates that the building's new energy system is deviating from the normal range and that there are abnormal phenomena in the building's new energy system.

[0015] Furthermore, the aforementioned building new energy system refers to the integration of one or more renewable energy technologies, such as photovoltaic power generation, solar thermal power generation, and ground source heat pumps, into a building, and the realization of diversified energy supply for power generation, heating, and cooling through intelligent control, forming a clean, low-carbon, efficient, and circular energy utilization system.

[0016] Further, S101: Obtain historical monitoring sample data of the building's new energy system, wherein the historical monitoring sample data includes: Sample data of photovoltaic system, solar thermal system, ground source heat pump system, indoor temperature and humidity, weather forecast data, and sub-item measurement and monitoring data were collected at set intervals. The sample data of the photovoltaic system includes: photovoltaic power generation, energy storage power, energy storage system discharge power, grid-connected power, and building power supply. The sample data of the photothermal system includes: inlet water temperature, outlet water temperature, system flow rate, inlet water pressure, and outlet water pressure; The sample data of the ground source heat pump system includes: inlet water temperature, outlet water temperature, and system flow rate; The indoor temperature and humidity sample data includes: the temperature and humidity inside the room; The weather forecast data includes: weather temperature, weather humidity, wind speed, wind direction, and solar radiation intensity; The sub-item monitoring data includes: power consumption of the HVAC system, power consumption of the lighting system, and power consumption of the power system.

[0017] For example, a monitoring system deployed on a photovoltaic system collects photovoltaic power generation data at certain time intervals. Energy storage capacity Discharge power of energy storage system Power output Building power supply ; The inlet water temperature is collected at regular intervals by a monitoring system deployed on the solar thermal system. Outlet water temperature System traffic Inlet water pressure Water pressure ; The inlet water temperature is collected at certain time intervals by a monitoring system deployed on the ground source heat pump system. Outlet water temperature System traffic .

[0018] Indoor temperature and humidity are collected at regular intervals by temperature and humidity sensors deployed in the main functional rooms. ; The local temperature and humidity are obtained at certain time intervals through the weather forecast system. Wind speed ,wind direction Solar radiation intensity .

[0019] The power consumption of the building's HVAC system at certain time intervals is obtained through a sub-item metering and monitoring system. Power consumption of lighting system Power consumption of the power system .

[0020] For example, a building energy consumption monitoring system acquires cumulative monitoring data for the temperature and humidity of main functional rooms, the solar photovoltaic system, and the building sub-metering system over the past year at 10-minute intervals. The monitoring data includes the power generation of the solar photovoltaic system, surplus electricity fed into the grid, the charging / discharging amount of the energy storage system, and the photovoltaic power generation consumed by the building sub-metering systems. A small weather station on the building rooftop acquires meteorological data such as ambient temperature and humidity, irradiance, wind speed, and wind direction.

[0021] Furthermore, after acquiring historical monitoring sample data of the building's new energy system and before preprocessing the historical monitoring sample data, the process further includes: The sample data of the photovoltaic system collected at set time intervals are used to generate a time series matrix of the photovoltaic system; each row of the matrix represents the time when the data was collected, and each column of the matrix represents the type of data; The sample data of the photothermal system collected at set time intervals are used to generate a time series matrix of the photothermal system; The sample data of the ground source heat pump system collected at set time intervals are used to generate a time series matrix of the ground source heat pump system. Indoor temperature and humidity sample data collected at set time intervals are used to generate an indoor temperature and humidity time series matrix. The weather forecast data collected at set intervals is used to generate a weather forecast time series matrix; The sub-item measurement and monitoring data collected at set time intervals are used to generate a sub-item measurement and testing time series matrix.

[0022] For example, after acquiring historical monitoring sample data of the building's new energy system and before preprocessing the historical monitoring sample data, the method further includes: By collecting photovoltaic (PV) operation data at certain time intervals through monitoring systems deployed on PV systems, a time series matrix of the PV system is formed. ; By collecting operational data of the solar thermal system at certain time intervals through a monitoring system deployed on the system, a time series matrix of the solar thermal system is formed. ; By collecting heat pump system operation data at certain time intervals through a monitoring system deployed on the heat pump system, a time series matrix of the heat pump system is formed. ; Indoor temperature and humidity are collected at regular intervals by temperature and humidity sensors deployed in the main functional rooms, forming an indoor temperature and humidity time series matrix. ; Meteorological parameters at specific time intervals are obtained from the weather forecasting system, including ambient temperature and humidity, radiation intensity, wind speed, etc., to form a time series matrix of meteorological parameters. ; By acquiring the energy consumption of different sub-systems of a building over a certain time interval through a building sub-metering system, a time series matrix of sub-system energy consumption is formed. .

[0023] Preprocessed data such as photovoltaic power generation and consumption are used to form a photovoltaic system time series matrix M. PV The preprocessed indoor temperature and humidity data of the main functional rooms are used to form an indoor temperature and humidity time series matrix M. T The preprocessed data, including ambient temperature and humidity, irradiance, wind speed, and wind direction, are used to form a meteorological parameter time series matrix M. Wp .

[0024] Furthermore, the historical monitoring sample data is preprocessed, including: using the k-means clustering algorithm to identify abnormal data, and using the kNN algorithm or kNN-Slope algorithm to repair abnormal data.

[0025] Further, S102: Construct a prediction model, which is implemented using a Transformer model.

[0026] The Transformer model is used for: (1-1): The preprocessed historical monitoring sample data is divided into several subsequences; (1-2): Each subsequence is treated as a sequence block. Each sequence block is input into the encoder of the Transformer model. The encoder encodes the sequence through a multi-head self-attention layer and a feedforward neural network layer. The encoder outputs the front-to-back relational representation of each sequence block. (1-3): The encoder inputs the output value into the decoder, and the decoder generates the predicted value within the set time range.

[0027] Further, in step (1-1): the preprocessed historical monitoring sample data is divided into several subsequences, including: (1-1-1): Establishing a multivariate time series matrix for building new energy systems ,in M It is the number of variables. L It refers to the number of time steps, with the goal of predicting the future. T The value of each time step ; (1-1-2): Divide the time series of each variable into N sequence blocks, each with a length of P and a step size of S. For the _th ... m There are 1 variable, including:

[0028] in, It is the first m The first variable n There are sequence blocks that satisfy: .

[0029] Further, in (1-2): each subsequence is treated as a sequence block, and each sequence block is input into the encoder of the Transformer model. The encoder encodes the sequence through a multi-head self-attention layer and a feedforward neural network layer. The encoder outputs a front-to-back relational representation of each sequence block, including: (1-2-1): Each sequence block is passed through a trainable linear mapping. and a position encoding matrix This is converted into the input notation of the Transformer, for the first... m There are variables that satisfy:

[0030] in, It is the input label of the nth sequence block of the mth variable, satisfying:

[0031] Represents the position encoding matrix The n A column vector, where the colon (:) indicates that all row elements of the column are retrieved.

[0032] (1-2-2): Concatenate the input labels of all variables to obtain a total input label matrix. ,satisfy:

[0033] (1-2-3): will The input is fed into a standard Transformer encoder to obtain an output tag matrix. ,satisfy: .

[0034] Further, in (1-3): the encoder inputs the output value into the decoder, and the decoder generates predicted values ​​within a set time range, including: Pass Z through a linear layer To obtain the final predicted value ,satisfy:

[0035] Furthermore, the prediction model is trained using preprocessed historical monitoring sample data to obtain the trained prediction model. The training process includes: The preprocessed historical monitoring sample data is input into the prediction model to train the prediction model. Training is stopped when the loss function value of the prediction model no longer decreases, or when the number of iterations exceeds the set number, and the trained prediction model is obtained.

[0036] The loss function of the prediction model adopts the cross-entropy loss function and the mean absolute error loss function to calculate the stage-by-stage deviation between the model prediction results and the actual monitoring results, and judges the stage-by-stage operation status of the building new energy system based on the average value.

[0037] Further, step S103: acquiring various monitoring variables of the target building, inputting these variables into the trained prediction model, and obtaining the prediction results of the various monitoring variables of the target building within a set prediction time range, including: Using the trained prediction model, calculate each variable t To t+ T The predicted value at time point is calculated, and its confidence interval is used as the normal threshold for the variable at the corresponding time point.

[0038] Further, in S104: first, a basic threshold is calculated, and then a dynamic threshold is determined based on the basic threshold, equipment aging factors, and seasonal factors; the prediction result is compared with the dynamic threshold, including: Calculate the base threshold ， Then, based on equipment aging and seasonal factors, dynamic thresholds are determined, and anomaly detection is performed, specifically including: First, the basic threshold is calculated using the sliding window percentile method based on historical MAPE. : ; For time window The set of historical mean absolute percentage error (MAPE) values ​​within the range, such as If so, then take the MAPE data from the past 30 days; The selected percentile function is [function name], and the selected percentile here is [percentage]. , which represents the upper limit of the allowable prediction error.

[0039] Calculate the aging coefficient of equipment : ; in, It refers to the service life; It refers to the number of repairs. It represents the number of failures; It is the ratio of the calculated service life to the set equipment lifespan (normalization of service life). It calculates the ratio of the current number of repairs to the set maximum number of repairs (normalization of the number of repairs); It is the ratio of the current number of failures to the set maximum number of failures (normalization of the number of failures); , , The weights of each indicator are determined by combining the Analytic Hierarchy Process (AHP) with expert experience. First, domain experts (such as equipment maintenance engineers and energy system designers) compare the impact of service life, number of maintenance, and number of failures in pairs to form a matrix. Then, the weights are solved by the eigenvalue method or geometric mean method.

[0040] Calculate the seasonality coefficient :

[0041] in, For the first Regression coefficients of meteorological parameters; For the first The weights of each parameter (determined through the Analytic Hierarchy Process (AHP) combined with expert experience); This represents the number of meteorological parameters.

[0042] calculate :

[0043] The weights of each indicator are set as follows: , , (Determined through the Analytic Hierarchy Process (AHP) combined with expert experience); The standardized value of the base threshold, This is the standardized value of the equipment aging factor. The standardized value of the seasonality coefficient is calculated using the following formula:

[0044]

[0045]

[0046] in, Based on the threshold, This is the maximum value of the basic threshold in historical data; This refers to the equipment aging factor. Equipment aging coefficient in historical data The maximum value; This is a seasonality coefficient. This represents the maximum value of the seasonality coefficient β in the historical data.

[0047] Calculate dynamic threshold :

[0048] in, Basic threshold; Standardized values ​​representing equipment aging information; Standardized values ​​representing the effects of seasonality; To adjust the coefficients and ensure Its effect is effective in the range of 0 to 1.

[0049] Perform anomaly detection: Within each detection window, a moving average is calculated based on the historical data of that window. The formula is:

[0050] Where N is the window length; The calculated With dynamic threshold Compare, if This indicates that the current prediction error is within the allowable range and the system is operating normally; if If the operating status deviates from the normal range, it indicates that there is energy efficiency degradation or abnormal phenomenon.

[0051] This invention relates to the field of efficient operation technology for building renewable energy systems. By integrating dynamic data monitoring information and combining multimodal anomaly data repair and dynamic threshold strategies, a closed-loop prediction and diagnosis system for building renewable energy systems is achieved. The system includes the following steps: acquiring historical fine-grained time-series monitoring sample data of the building renewable energy system and building sub-metering, and processing outliers and filling in missing values ​​in the sample data; constructing and training a Transformer-based multivariate medium- to long-term prediction model for the building renewable energy system; using the trained model to synchronously predict various monitoring variables of the target building; and based on the synchronous prediction results and monitoring results, dynamically assessing the degree of system energy efficiency degradation by defining a Composite Health Index (CHI), and triggering an operational status warning when the prediction deviation exceeds a threshold. This invention utilizes an attention-based Transformer model to achieve multivariate medium- to long-term synchronous prediction, and uses dynamic confidence intervals and the CHI to accurately diagnose the system's operational status, thereby supporting the timely triggering of optimized control strategies when abnormal states occur, ensuring the efficient and safe operation of the building renewable energy system.

[0052] Example 2 This embodiment provides a device for predicting the operational status of a building's new energy system, including: The acquisition module is configured to: acquire historical monitoring sample data of the building's new energy system and preprocess the historical monitoring sample data; The training module is configured to: build a prediction model, train the prediction model using preprocessed historical monitoring sample data, and obtain the trained prediction model. The prediction module is configured to: acquire various monitoring variables of the target building, input the various monitoring variables of the target building into the trained prediction model, and obtain the prediction results of the various monitoring variables of the target building within the set prediction time range; The status assessment module is configured to: assess the operational status of the target building based on the prediction results; first, calculate the basic threshold, and then determine the dynamic threshold based on the basic threshold, equipment aging factors, and seasonal factors; compare the prediction results with the dynamic thresholds; if the prediction results are less than or equal to the dynamic thresholds, it indicates that the building's new energy system is operating normally; if the prediction results are greater than the dynamic thresholds, it indicates that the building's new energy system is deviating from the normal range and that there are abnormal phenomena in the building's new energy system.

[0053] It should be noted that the acquisition module, training module, prediction module, and state evaluation module described above correspond to steps S101 to S104 in Embodiment 1. The examples and application scenarios implemented by these modules and their corresponding steps are the same, but they are not limited to the content disclosed in Embodiment 1. It should also be noted that these modules, as part of the system, can be executed in a computer system, such as a set of computer-executable instructions.

[0054] The descriptions of each embodiment in the above embodiments have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0055] The proposed system can be implemented in other ways. For example, the system embodiments described above are merely illustrative, and the division of modules described above is only a logical functional division. In actual implementation, there may be other division methods. For example, multiple modules may be combined or integrated into another system, or some features may be ignored or not executed.

[0056] Example 3 This embodiment also provides an electronic device, including: one or more processors, one or more memories, and one or more computer programs; wherein, the processor is connected to the memory, and the one or more computer programs are stored in the memory. When the electronic device is running, the processor executes the one or more computer programs stored in the memory to cause the electronic device to perform the method described in Embodiment 1.

[0057] It should be understood that in this embodiment, the processor can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc.

[0058] Memory may include read-only memory and random access memory, and provides instructions and data to the processor. A portion of memory may also include non-volatile random access memory. For example, memory may also store information about the device type.

[0059] In the implementation process, each step of the above method can be completed by the integrated logic circuits in the processor hardware or by software instructions.

[0060] The method in Embodiment 1 can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor. The software modules can reside in readily available storage media in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory; the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method. To avoid repetition, a detailed description is not provided here.

[0061] Those skilled in the art will recognize that the units and algorithm steps described in connection with the various examples of this embodiment can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention.

[0062] Example 4 This embodiment also provides a computer-readable storage medium for storing computer instructions, which, when executed by a processor, complete the method described in Embodiment 1.

[0063] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for predicting the operational status of building renewable energy systems, characterized by: include: Acquire historical monitoring sample data of building new energy systems and preprocess the historical monitoring sample data; A prediction model is constructed, and the prediction model is trained using preprocessed historical monitoring sample data to obtain the trained prediction model. Obtain the various monitoring variables of the target building, input the various monitoring variables of the target building into the trained prediction model, and obtain the prediction results of the various monitoring variables of the target building within the set prediction time range; Based on the prediction results, the operating status of the target building is evaluated: first, a basic threshold is calculated, and then a dynamic threshold is determined based on the basic threshold, equipment aging factors, and seasonal factors; the prediction results are compared with the dynamic threshold. If the prediction results are less than or equal to the dynamic threshold, it indicates that the operating status of the building's new energy system is normal; if the prediction results are greater than the dynamic threshold, it indicates that the operating status of the building's new energy system deviates from the normal range, and there is an abnormal phenomenon in the building's new energy system.

2. The method for predicting the operational status of a building renewable energy system as described in claim 1, characterized in that, Obtain historical monitoring sample data of building new energy systems, including: Sample data of photovoltaic system, solar thermal system, ground source heat pump system, indoor temperature and humidity, weather forecast data, and sub-item measurement and monitoring data were collected at set intervals. The sample data of the photovoltaic system includes: photovoltaic power generation, energy storage power, energy storage system discharge power, grid-connected power, and building power supply. The sample data of the photothermal system includes: inlet water temperature, outlet water temperature, system flow rate, inlet water pressure, and outlet water pressure; The sample data of the ground source heat pump system includes: inlet water temperature, outlet water temperature, and system flow rate; The indoor temperature and humidity sample data includes: the temperature and humidity inside the room; The weather forecast data includes: weather temperature, weather humidity, wind speed, wind direction, and solar radiation intensity; The sub-item monitoring data includes: power consumption of the HVAC system, power consumption of the lighting system, and power consumption of the power system.

3. The method for predicting the operational status of a building renewable energy system as described in claim 1, characterized in that, After acquiring historical monitoring sample data of the building's new energy system, and before preprocessing the historical monitoring sample data, the process further includes: The sample data of the photovoltaic system collected at set time intervals are used to generate a time series matrix of the photovoltaic system; each row of the matrix represents the time when the data was collected, and each column of the matrix represents the type of data; The sample data of the photothermal system collected at set time intervals are used to generate a time series matrix of the photothermal system; The sample data of the ground source heat pump system collected at set time intervals are used to generate a time series matrix of the ground source heat pump system. Indoor temperature and humidity sample data collected at set time intervals are used to generate an indoor temperature and humidity time series matrix. The weather forecast data collected at set intervals is used to generate a weather forecast time series matrix; The sub-item measurement and monitoring data collected at set time intervals are used to generate a sub-item measurement and testing time series matrix.

4. The method for predicting the operational status of a building renewable energy system as described in claim 1, characterized in that, A prediction model is constructed, which is implemented using the Transformer model; The Transformer model is used for: (1-1): The preprocessed historical monitoring sample data is divided into several subsequences; (1-2): Each subsequence is treated as a sequence block. Each sequence block is input into the encoder of the Transformer model. The encoder encodes the sequence through a multi-head self-attention layer and a feedforward neural network layer. The encoder outputs the front-to-back relational representation of each sequence block. (1-3): The encoder inputs the output value into the decoder, and the decoder generates the predicted value within the set time range.

5. The method for predicting the operational status of a building renewable energy system as described in claim 4, characterized in that, The step (1-1) involves dividing the preprocessed historical monitoring sample data into several sub-sequences, including: (1-1-1): Establishing a multivariate time series matrix for building new energy systems ,in M It is the number of variables. L It refers to the number of time steps, with the goal of predicting the future. T The value of each time step ; (1-1-2): Divide the time series of each variable into N sequence blocks, each with a length of P and a step size of S. For the _th ... m There are 1 variable, namely: ； in, It is the first m The first variable n There are sequence blocks that satisfy: 。 6. The method for predicting the operational status of a building renewable energy system as described in claim 4, characterized in that, (1-2): Each subsequence is treated as a sequence block, and each sequence block is input into the encoder of the Transformer model. The encoder encodes data through a multi-head self-attention layer and a feedforward neural network layer. The encoder outputs a relational representation of the preceding and following sequences for each sequence block, including: (1-2-1): Each sequence block is passed through a trainable linear mapping. and a position encoding matrix This is converted into the input notation of the Transformer, for the first... m There are variables that satisfy: ； in, It is the input label of the nth sequence block of the mth variable, satisfying: ； Represents the position encoding matrix The n A column vector, where the colon (:) indicates that all row elements of the column are retrieved; (1-2-2): Concatenate the input labels of all variables to obtain a total input label matrix. ,satisfy: ； (1-2-3): will The input is fed into a standard Transformer encoder to obtain an output tag matrix. ,satisfy: ； (1-3): The encoder inputs the output value into the decoder, and the decoder generates predicted values ​​within a set time range, including: Pass Z through a linear layer To obtain the final predicted value ,satisfy: 。 7. The method for predicting the operational status of a building renewable energy system as described in claim 1, characterized in that, First, calculate the baseline threshold, then determine the dynamic threshold based on the baseline threshold, equipment aging factors, and seasonal factors; compare the predicted results with the dynamic threshold, including: Calculate the base threshold ， Then, based on equipment aging and seasonal factors, dynamic thresholds are determined, and anomaly detection is performed, specifically including: First, the basic threshold is calculated using the sliding window percentile method based on historical MAPE. : ； For time window The set of historical mean absolute percentage error (MAPE) values ​​within the range, The selected percentile function is [function name], and the selected percentile here is [percentage]. , which represents the upper limit of the allowable prediction error; Calculate the aging coefficient of equipment : ； in, It refers to the service life; It refers to the number of repairs. It represents the number of failures; It is the ratio of the calculated service life to the set equipment lifespan; It is the ratio of the current number of repairs to the set maximum number of repairs; It is the ratio of the current number of failures to the set maximum number of failures; , , The weights of each indicator are determined by combining the analytic hierarchy process with expert experience. First, domain experts compare the impact of service life, number of maintenance, and number of failures pairwise to form a matrix. Then, the weights are solved by the eigenvalue method or the geometric mean method. Calculate the seasonality coefficient : ； in, For the first Regression coefficients of meteorological parameters; For the first The weights of each parameter; The number of meteorological parameters; calculate : ； The weights of each indicator are set as follows: , , ; The standardized value of the base threshold, This is the standardized value of the equipment aging factor. The standardized value of the seasonality coefficient is calculated using the following formula: ； ； ； in, Based on the threshold, This is the maximum value of the basic threshold in historical data; The aging factor of the equipment. Equipment aging coefficient in historical data The maximum value; This is a seasonality coefficient. This represents the maximum value of the seasonality coefficient β in historical data; Calculate dynamic threshold : ； in, Basic threshold; Standardized values ​​representing equipment aging information; Standardized values ​​representing the effects of seasonality; To adjust the coefficients and ensure The effect is effective in the range of 0 to 1; Anomaly detection: Within each detection window, a moving average is calculated based on the historical data of that window. The formula is: ； Where N is the window length; The calculated With dynamic threshold Compare, if This indicates that the current prediction error is within the allowable range and the system is operating normally; if If the operating status deviates from the normal range, it indicates that there is energy efficiency degradation or abnormal phenomenon.

8. A device for predicting the operational status of a building's new energy system, characterized in that, include: The acquisition module is configured to: acquire historical monitoring sample data of the building's new energy system and preprocess the historical monitoring sample data; The training module is configured to: build a prediction model, train the prediction model using preprocessed historical monitoring sample data, and obtain the trained prediction model. The prediction module is configured to: acquire various monitoring variables of the target building, input the various monitoring variables of the target building into the trained prediction model, and obtain the prediction results of the various monitoring variables of the target building within the set prediction time range; The status assessment module is configured to: assess the operational status of the target building based on the prediction results; first, calculate the basic threshold, and then determine the dynamic threshold based on the basic threshold, equipment aging factors, and seasonal factors; compare the prediction results with the dynamic thresholds; if the prediction results are less than or equal to the dynamic thresholds, it indicates that the building's new energy system is operating normally; if the prediction results are greater than the dynamic thresholds, it indicates that the building's new energy system is deviating from the normal range and that there are abnormal phenomena in the building's new energy system.

9. An electronic device, characterized in that it comprises: Memory is used to store computer-readable instructions in a non-transitory manner. as well as Processor, for executing the computer-readable instructions, When the computer-readable instructions are executed by the processor, they perform the method described in any one of claims 1-7.

10. A storage medium, characterized in that, Non-transitory storage of computer-readable instructions, wherein when the non-transitory computer-readable instructions are executed by a computer, the method of any one of claims 1-7 is performed.

Images